Parathyroid Hormone Enhances Early and Suppresses Late Stages of Osteogenic and Chondrogenic Development in a BMP-Dependent Mesenchymal Differentiation System (C3H10T½)

Authors

  • Angela Hollnagel,

    1. Gesellschaft für Biotechnologische Forschung (GBF), “Growth Factors and Receptors,” Braunschweig, Germany
    Search for more papers by this author
  • Marion Ahrens,

    1. Gesellschaft für Biotechnologische Forschung (GBF), “Growth Factors and Receptors,” Braunschweig, Germany
    Search for more papers by this author
  • Gerhard Gross

    Corresponding author
    1. Gesellschaft für Biotechnologische Forschung (GBF), “Growth Factors and Receptors,” Braunschweig, Germany
    • Address reprint requests to: Gerhard Gross, Gesellschaft für Biotechnologische, Forschung, Mascheroder Weg 1, 38124 Braunschweig, Germany
    Search for more papers by this author

  • This investigation was presented at the XVth Meeting of the Federation of the European Connective Tissue Societies 1996 in Munich, Germany (abstract).

Abstract

The role of parathyroid hormone (PTH) upon osteo-/chondrogenic development was investigated in a bone morphogenetic protein (BMP)-dependent differentiation system involving the recombinant expression of BMPs in mesenchymal progenitor cells (C3H10T½). The constitutive expression of the PTH/PTH related protein receptor in this system led to a marked stimulation of chondrogenic and osteogenic development, while the permanent application of the ligand PTH(1–34) resulted in opposite responses by stimulating the early and suppressing the late stages of osteo-/chondrogenic development. These contrasting effects of PTH(1–34) on osteogenic and chondrocytic development seem, therefore, to depend on the cellular state of differentiation. The osteogenic and chondrocytic differentiation potential was substantiated histologically and by genetic analyses of marker genes like c-fos, alkaline phosphatase, osteocalcin, collagen α1(I), and collagen α1(II). The capacity to regulate osteogenic and chondrogenic development is located in the amino-terminal (1–34) region of the PTH molecule and seems to be mediated by the cyclic adenosine monophosphate signaling cascade. The application of other PTH domains like PTH(28–48) and PTH(53–84) did not exhibit significant responses. PTH acts as an essential factor in mesenchymal development controlling rates of differentiation into the osteogenic or chondrogenic lineage. The analysis of PTH effects in this system demonstrates the value of recombinant mesenchymal progenitor cells in the in vitro analysis of osteo-/chondrogenic development.

INTRODUCTION

PARATHYROID HORMONE (PTH) is secreted from the parathyroid gland and regulates calcium and phosphate transport in the kidney and deeply influences bone synthesis and remodeling.1,2 In contrast, the PTH related protein (PTHrP) is expressed in a wide variety of tissues and seems to act in a paracrine and autocrine fashion.3,4 PTHrP was first identified causing malignancy-associated hypercalcemia.5,6 The targeted disruption of PTHrP in mice is postnatally lethal and displays many abnormalities in endochondral bone formation.7 A similar, but more severe, phenotype that may lead to death in utero is observed in mice that lack the PTH/PTHrP receptor.8 PTHrP has very recently been shown to play a role by limiting the number of proliferating chondrocytes that enter the hypertrophic differentiation process.9

The PTH/PTHrP receptor has been cloned and characterized as a G-coupled receptor containing seven transmembrane domains.10,11 PTHrP and PTH bind this receptor on the surface of osteoblastic, chondrocytic, and renal epithelial cells specifically through a conserved 34 amino acid amino-terminal region. Binding of either ligand leads to an increase in intracellular cyclic adenosine monophosphate (cAMP) synthesis, raises the intracellular level of calcium, and stimulates inositol 1,4,5-trisphosphate production.11 During embryonic development, the gene encoding the PTH/PTHrP receptor is expressed early on and seems to be involved in extraembryonic endoderm development. It is expressed at many sites of mesenchymal/epithelial interactions and, at later stages, in various organs and growth plate chondrocytes, suggesting a profound role in embryonic development and bone formation.12

The PTH-mediated effects on bone growth in the adult organism are complex since PTH provokes catabolic as well as anabolic responses. PTH, administered intermittently for short time periods, is catabolic for bone.13–15 However, if administered over prolonged periods, it exerts anabolic effects in normal and osteoporotic animals or humans.16–18 The precise mechanism of PTH action in vitro appears controversial, since PTH either stimulates or suppresses the growth of osteoblasts and chondrocytes and exerts differentiating as well as dedifferentiating effects on the osteoblastic phenotype.19,20 The amino-terminal PTH fragment PTH(1–34) is a ligand for the PTH/PTHrP receptor and, in vivo, is necessary and sufficient to induce the various PTH-specific calcemic effects. A potential biological function of the midregional and the carboxy-terminal PTH fragments is a matter of numerous studies (also see Discussion).

Osteoblasts and chondrocytes are part of the system that is involved in bone formation (e.g., reviewed in Ref. 21). The exact genealogy leading to differentiation of mesenchymal cells to develop from progenitors into osteoblasts or chondrocytes is not absolutely established. However, bone morphogenetic proteins (BMPs) have been isolated from adult bone on the basis of their ability to induce a cascade of events leading to ectopic bone formation if implanted subcutaneously.22 These implants recapitulate a sequence of events including chemotaxis, cellular proliferation, and differentiation closely resembling embryonic long bone development.23 In adult life, BMPs may be part of the complex system that regulates the maintenance of mass and mineral in bone and cartilage. BMPs belong to the transforming growth factor β (TGF-β) superfamily and share a high identity to other closely related proteins which have been characterized in vertebrates and invertebrates to be involved in embryonic development, body patterning, and mesoderm specification (reviewed by Hogan).24

An increase in PTH/PTHrP receptor levels is a hallmark during the osteoblast developmental sequence. The latter phenomenon has also been demonstrated in mesenchymal progenitor cells (C3H10T½) where exogenous application or vector-mediated constitutive expression of BMP2 or BMP4 is sufficient to determine in vitro the development into three mesenchymal lineages: the chondrogenic, the osteogenic, and the adipogenic lineage.25–27 With this system, we examined (1) how enhanced levels of the PTH/PTHrP receptor and (2) its ligand represented by various PTH domains (the amino-terminal 1–34; the midregional 28–48; and the carboxy-terminal 53–84 fragment) influence this BMP2- or BMP4-dependent development. We find that overexpression of the PTH/PTHrP receptor in BMP-expressing C3H10T½ cells dramatically enhance chondrogenic and osteogenic development, while the addition of the ligand PTH(1–34) leads to differentiation state–specific opposite responses in stimulating the early and suppressing the late stages of osteogenic or chondrogenic development.

MATERIALS AND METHODS

Cell lines, culture conditions, and transfection experiments

The features of BMP2- and BMP4-transfected C3H10T½ cells have been described by Ahrens et al.25 Human BMP2 and BMP4 are constitutively expressed under the control of the LTR of the myeloproliferative sarcoma virus (MPSV).25,28 If not stated otherwise, cells were plated at a density of 5000 cells/cm2. Cells were routinely grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal calf serum. PTH/PTHrP receptor was cloned from ROS17/2.8 cells using the published sequence information.11 Expression of the PTH/PTHrP receptor was also under the control of the LTR from MPSV. Transfection was performed by calcium phosphate precipitation. Control or BMP-transfected C3H10T½ cells were selected by cotransfection with a plasmid mediating resistance against puromycin (5 μg/ml). Puromycin-resistant colonies were subcultivated, and selection pressure was maintained during the entire cultivation period to follow. C3H10T½ cells harboring the rat PTH/PTHrP receptor were selected by cotransfection with the plasmid-mediating resistance against G418 (750 μg/ml). Cells were maintained in this concentration of G418. After reaching confluence (arbitrarily termed day 0), 50 μg/ml ascorbic acid and 10 mM β-glycerophosphate were added as specified for the cultivation of native osteoblast-like cells.29 All experiments were performed as well in C3H10T½BMP2 as in C3H10T½BMP4 cells. The outcome of these experiments was essentially identical and independent of the BMP background. Therefore, all analyses will be described with C3H10T½BMP2 and C3H10T½BMP2-PTHR cells only.

PTH treatment

PTH(1–34), PTH(28–48), and PTH(53–84) were synthesized chemically by the peptide synthesis unit of the GBF and were added at the time of plating at a concentration of 1 × 10−7 M. The medium with or without the PTH fragments was routinely replaced every 2 days of cultivation.

Adenylate cyclase stimulation and cAMP determination

Cells were cultured in 48-well plates (12,500 cells/cm2) as outlined above. After 6 days of cultivation, cells were washed twice with serum-free medium and preincubated for 20 minutes at 37°C in serum-free medium containing 1 mg/ml bovine serum albumin (BSA) and 1 mM isobutylmethylxanthine (IBMX; Sigma Chemical Co., St. Louis, MO, U.S.A.), which inhibits phosphodiesterase activity. Stimulation was performed by an additional 15-minute incubation with the corresponding PTH fragment. The incubation was stopped by aspirating the medium and washing the cells once with ice-cold phosphate-buffered saline (PBS). Incubations were conducted in triplicate. cAMP was extracted with 1 ml of acidified ethanol (1.75 ml HCl concentrated/100 ml ethanol) overnight at −20°C. The extract was removed, evaporated to dryness, and redissolved in 200 μl of PBS. Duplicates were assayed for cAMP by a competitive binding method using a commercially available [3H]cAMP radioassay kit (Amersham Buchler). If not stated otherwise, experimental values are expressed as picomoles of cAMP per well.

Thymidine incorporation assay

The rate of DNA synthesis was assayed by the incorporation of [3H]thymidine into perchloric acid–precipitable material. Cells were cultivated in 200 μl of medium for 30 h in 96-well microtiter plates as described.30

Statistics

All data were obtained in duplicate or triplicate in at least three independent experiments. Data are presented in a quantitative fashion as the mean and SEM.

mRNA analysis

C3H10T½ cells harboring the expression vectors were cultivated as described above. Cells were harvested at the indicated time intervals, and total RNA was isolated by guanidinium/CsCl step gradients.

Northern blot analysis:

Total cellular RNA (10 μg) was separated electrophoretically in a 2.2 M formaldehyde −1.2% agarose gel and transferred to nitrocellulose. Hybridization was carried out with nick-translated32P-labeled gene-specific DNA probes. mRNA levels were related to rRNA levels to correct for loading variations. Quantification was performed with an imaging system (WinCam 2.0, Cybertech, Berlin, Germany).

Reverse transcription polymerase chain reaction:

Total RNA (10 μg) was reverse transcribed after annealing with 2.5 ng random hexamer primers (GIBCO-BRL, Grand Island, NY, U.S.A.) for 2 minutes at 56°C using 10 U SuperScript RNAse H-Reverse Transcriptase (GIBCO-BRL) in a 50 μl reaction mix containing 50 mM Tris, pH 8.3, 75 mM KCl, 3 mM MgCl2, 10 mM dithiothreitol, 0.5 mM dATP, dCTP, dGTP, dTTP each (Pharmacia, Uppsala, Sweden), 40 U RNAsin (Promega Biotech Co., Madison, WI, U.S.A.) for 1 h at 37°C. One microliter of the single-stranded cDNA sample was used in a 25 μl reaction containing 10 mM Tris, pH 9.0, 50 mM KCl, 0.1% Triton X–100, 2.5 mM MgCl2, 0.4 mM dATP, dCTP, dGTP, dTTP, 20 pmol of osteocalcin- and β-actin gene-specific primers (osteocalcin: GCAGACCTAGCAGACACCAT and GAGCTGCTGTGACATCCATAC; β-actin: TCTACAATGAGCTGCGTGTGG and AGTACTTGCGCTCAGGAGG, respectively), and 2.5 U Taq DNA polymerase (Promega). The following program was used: one cycle with 7 minutes of denaturation at 96°C, 2 minutes of annealing at 49°C (hot start), followed by 30 cycles of denaturation at 96°C for 45 s, annealing at 49°C (25 s), and extension at 73°C (3 minutes).

Histological methods and verification of cellular phenotypes

Osteoblasts exhibit a stellate morphology displaying high levels of alkaline phosphatase (ALP) activity which was visualized by cellular staining with α-naphthyl-phosphate and Fast Red. Deposition of mineral was monitored by the von Kossa staining method.29 Chondrocytes were identified by staining with Alcian Blue at pH 2.5. They also displayed ALP activity but in comparison with osteoblasts of lower intensity and they exhibited a round cell morphology.

RESULTS

Expression profile of PTH/PTHrP receptors

The recombinant stable expression of the rat PTH/PTHrP receptor in murine C3H10T½ cells and in C3H10T½ cells expressing recombinant BMP2 (C3H10T½BMP2) or BMP4 was under the control of the MPSV-LTR (Materials and Methods). The outcome of these experiments was essentially identical and independent of the BMP type. Therefore, all analyses will be described in a BMP2 background, only. About 5000 transfectants were pooled, cultivated, and the level of expression during the entire cultivation period was monitored by analyzing the level of PTH/PTHrP receptor mRNA expression and by the determination of the ligand-mediated stimulation of cAMP synthesis. Cells expressing the recombinant rat PTH/PTHrP receptor show a high amount of PTH/PTHrP receptor mRNA in comparison with the parental C3H10T½BMP2 or C3H10T½ cells (Fig. 1a). In C3H10T½BMP2 cells, PTH/PTHrP receptor mRNA is monitored after considerably longer exposures only (Fig. 1a).

Figure FIG. 1.

PTH/PTHrP receptor expression in parental and recombinant C3H10T½ cells. (a) Northern blot analyses of C3H10T½ cells expressing recombinant BMP2 (C3H10T½BMP2) and/or the PTH/PTHrP receptor (C3H10T½BMP2-PTHR or C3H10T½PTHR, respectively). Cells were harvested at the indicated time intervals (days postconfluence) and 10 μg of total RNA was applied to Northern analyses. Autoradiographs are shown after 6 h of exposure; *2 week exposure. (b) Ligand-dependent cAMP synthesis. The biological activity of the PTH/PTHrP receptors was assessed by the PTH(1–34)-dependent activation (300 nM) of adenylate cyclase in comparison with the activation by forskolin (50 μM) and a control treatment (acetic acid; 10 mM). Values were determined at day 1 postconfluence and evaluated as described in Materials and Methods. Values represent the mean of at least three independent assays. (c) Ligand-dependent PTH/PTHrP receptor desensitization. Cells were grown in the permanent presence of the ligand PTH(1–34) (100 nM) (Materials and Methods) followed by the determination of ligand-dependent cAMP synthesis.

The stable transfection of PTH/PTHrP receptor encoding expression vectors results in a PTH(1–34)-dependent 9- to 170-fold stimulation of cAMP synthesis (Fig. 1b). This was in the range of forskolin-mediated cAMP synthesis, which is a direct stimulator of adenylate cyclase (Fig. 1b). This high level of PTH(1–34)-dependent cAMP synthesis was maintained throughout the entire cultivation period in C3H10T½BMP2-PTHR cells (Fig. 1c). To investigate whether or not C3H10T½BMP2 cells respond to continuous PTH(1–34) treatment by a down-regulation of ligand-dependent cAMP synthesis, PTH(1–34) (100 nM) was added to the medium with plating and then routinely replaced every 2 days. The level of receptor desensitization was assessed by the comparison of PTH(1–34)-induced cAMP synthesis in cells grown in the presence versus cells grown in the absence of PTH(1–34) (Fig. 1c). In C3H10T½BMP2 cells, the level of receptor desensitization was in the range of one order of magnitude (Fig. 1c). In C3H10T½BMP2-PTHR cells, this effect is severely reduced or even lost, indicating that a homologous down-regulation of PTH/PTHrP receptors does not take place in cells expressing high levels of recombinant PTH/PTHrP receptors.

The influence of PTH/PTHrP receptors on osteogenic and chondrogenic development in recombinant C3H10T½ cells

After reaching confluence, an extensive matrix production was observed for BMP-transfected (C3H10T½BMP2) and the BMP and PTH/PTHrP receptor–transfected C3H10T½ cells (C3H10T½BMP2-PTHR). These cells grow in multilayers, reaching a final cell density of ∼0.7 × 106 cells/cm2. In contrast, native C3H10T½ cells or cells that express recombinant PTH/PTHrP receptors in the absence of recombinant BMP2 (C3H10T½PTHR) ceased proliferation upon reaching confluence (final cell density, ∼0.2 × 106 cells/cm2). The amino-terminal, midregional, or carboxy-terminal PTH fragments in the cultivation medium cells did not significantly influence cell growth or proliferation, as exemplified also by thymidine incorporation assays (data not shown).

Histological analyses of osteogenic and chondrogenic development were performed 10–15 days postconfluence (Fig. 2). C3H10T½BMP2-PTHR showed more and larger chondrogenic as well as osteogenic clonal amplification centers in comparison with C3H10T½BMP2 cells (Fig. 2a, 2b, 2c, 2d, 2e). In addition, areas of beginning mineralization were identified in C3H10T½BMP2-PTHR cells only. Osteogenic development was also verified by Northern analyses. In C3H10T½BMP2-PTHR cells, mRNA levels of osteogenic marker genes are enhanced in comparison with C3H10T½BMP2 cells (Fig. 3). c-fos expression exhibits the typical biphasic nature of expression as it has been characterized for osteogenic development in primary cell systems.31,32 In addition, the striking 20- to ∼200-fold increase in collagen α1(II) mRNA levels in C3H10T½BMP2-PTHR in comparison with C3H10T½BMP2 cells indicates enhanced development into the chondrogenic lineage (Fig. 3). This is in accordance with the histological manifestation of large and extended centers of chondrogenic development at day 12 postconfluence (Fig. 2b).

Figure FIG. 2.

Histological analysis of osteogenic and chondrogenic development. (a) Alcian blue positive, chondrocyte-like amplification centers in C3H10T½BMP2 cells (12 days postconfluence). (b) Alcian blue positive, chondrocyte-like amplification centers in C3H10T½BMP2-PTHR cells (12 days postconfluence). (c) ALP positive, osteoblast-like amplification centers in C3H10T½BMP2 cells (10 days postconfluence) (d) ALP positive, osteoblast-like amplification centers in C3H10T½BMP2-PTHR cells (10 days postconfluence). (e) Roux flasks of C3H10T½BMP2 and C3H10T½BMP2-PTHR cells. Alcian Blue positive chondrogenic centers at day 12 (left) and ALP positive osteogenic centers at day 10 postconfluence (right). (f) Calcifying nodules at day 15 post-confluence in C3H10T½BMP2-PTHR cells visualized by histological staining for osteoblasts (alkaline phosphatase; red) and mineralization (van Kossa; black).

Figure FIG. 3.

Northern analysis of marker genes for osteoblast and chondrocyte development. The level of rRNA is shown to indicate loading variations. Numbers indicated the relative amount of mRNA as determined by an imaging system (Materials and Methods).

Influence of PTH fragments on osteogenic and chondrogenic development

PTH fragments (PTH(1–34), PTH(28–48), and PTH(53–84)) were added to the cultivation medium (100 nM) upon plating and were replaced with the medium every 2 days until the end of the cultivation period. Chondrocyte formation was investigated at day 12 postconfluence. Application of the ligand PTH(1–34) enhances chondrogenic development in C3H10T½BMP2 cells but suppresses chondrogenesis in C3H10T½BMP2-PTHR cells (Fig. 4). The inhibitory effect of the ligand PTH(1–34) upon rates of chondrocytic differentiation of C3H10T½BMP2-PTHR cells suggests that high levels of activated PTH/PTHrP receptor suppress chondrogenic development, and that modest levels of activated PTH/PTHrP-receptor as present in C3H10T½BMP2-PTHR cells or in PTH(1–34)–treated C3H10T½BMP2 cells are stimulatory for ongoing chondrogenesis (Fig. 1).

Figure FIG. 4.

Chondrocyte formation in recombinant C3H10T½BMP2 or C3H10T½BMP2-PTHR cells at day 12 postconfluence. Cells were continuously cultivated in the presence of PTH(1–34) (100 nM), forskolin (100 μM), or TPA (1 μM). Analyses were performed by assessing the number of Alcian Blue, chondrocyte-like cells present. Values represent mean values of a typical experiment in triplicates.

Regarding osteogenic development, the opposite stimulatory as well as inhibitory phenomenon is even more obvious (Figs. 5a and 5b). With C3H10T½BMP2-PTHR cells, continuous PTH(1–34) application leads to the fast and advanced formation of ALP positive, osteoblast-like clonal amplification centers which are visible already at very early stages of the cultivation (shortly before reaching confluence), whereas in the absence of the ligand these osteogenic amplification centers are only observed 3–4 days later (Fig. 5a). These centers in the untreated cells are able to enter the final stages of the osteoblast developmental sequence and can form calcifying nodules eventually (Fig. 2f). PTH(1–34) treatment, however, prevents the formation of these nodules and leads to a decrease of alkaline positive cells at later stages of cultivation (days 7–15). C3H10T½BMP2-PTHR cells assume an extended fibroblastic nature upon continuous PTH(1–34) treatment and ALP positive, cuboidal-shaped osteoblast-like cells or Alcian Blue stained, round chondrocytes are not observed.

Figure FIG. 5.

Influence of PTH(1–34) treatment on osteogenic and chondrogenic development. (a) Histological analysis of osteogenic development in the presence or absence of the ligand PTH(1–34) (100 nM) in C3H10T½BMP2-PTHR cells by ALP staining. (b) Formation of osteoblast-like cells in recombinant C3H10T½BMP2-PTHR cells. The cells were continuously cultivated in the presence of PTH(1–34) (100 nM), forskolin (100 μM), or TPA (1 μM). Histological analyses were performed at the indicated time intervals by assessing the number of cuboidal, ALP positive osteoblast-like cells. Values represent mean values of a typical experiment in triplicates.

In addition, Northern analyses of C3H10T½BMP2-PTHR cells (Fig. 6a) verify that continuous PTH(1–34) treatment decreases the expression of the major osteo-/chondrogenic marker genes, such as collagen α1 (II) (∼4-fold) and ALP (∼3- to 5-fold) in comparison with untreated cells (Fig. 3). The level of collagen α1 (I) mRNA is also reduced at late stages of the cultivation (∼2-fold), and, importantly, the enhanced expression of the osteoblast-specific osteocalcin mRNA at late stages of osteogenic development (days 10–15) is absent in PTH(1–34)-treated C3H10T½BMP2-PTHR cells in comparison with untreated cells. The suppression of osteocalcin mRNA levels in PTH(1–34)-treated C3H10T½BMP2-PTHR cells was also substantiated by reverse transcribed polymerase chain reaction (RT-PCR; Fig. 6b). These results indicate that active PTH/PTHrP receptors interfere with the late stages of the osteoblast developmental sequence where osteocalcin is up-regulated. Continuous PTH(1–34) treatment of C3H10T½BMP2-PTHR leads to increased c-fos mRNA levels (Fig. 6a), suggesting that elevated levels of c-fos expression do not necessarily contribute to osteo-/chondrogenic development, especially at late stages of development.

Figure FIG. 6.

(a) Osteogenic and chondrogenic marker gene expression in C3H10T½BMP2-PTHR cells in the continuous presence of PTH(1–34) (100 nM). Northern analyses were performed as described in Fig. 1a and Materials and Methods. The level of mRNA is shown to indicate loading variations. The relative amount of mRNA is indicated as determined by an imaging system (Materials and Methods). (b) PTH(1–34)-dependent osteocalcin mRNA expression in recombinant C3H10T½ cells analyzed by RT-PCR (Materials and Methods). The control RT-PCR (β-actin) yields identical results among the mRNAs investigated and is shown as obtained for mRNA of PTH(1–34)-treated C3H10T½BMP2-PTHR cells.

In none of our investigations did we observe an influence of the midregional (PTH(28–48)) or carboxy-terminal fragments (PTH(53–84)) on rates of mesenchymal differentiation in morphological as well as in genetic terms (data not shown).

Influence of forskolin or TPA on osteogenic and chondrogenic development

To characterize further the stimulatory or inhibitory action of PTH, forskolin (100 μM) or 12-O-tetradecanoylphorbol–13-acetate (TPA; 1 μM) was continuously added to the cultures. Forskolin was capable of mimicking PTH(1–34) activities in C3H10T½BMP2-PTHR cells; ALP positive osteoblast-like cells were detected at early forskolin-treated stages of cultivation (Fig. 5b). Moreover, in C3H10T½BMP2 cells (Fig. 5b) as well as in the parental C3H10T½ and in C3H10T½PTHR cells, a forskolin-dependent increase in ALP positive cells was monitored (data not shown). But as with PTH(1–34), the permanent presence of forskolin interfered with the morphological manifestation of the osteogenic and chondrogenic phenotype at late stages of cultivation. TPA, a direct stimulator of the PKC pathway, did not exhibit this feature (Fig. 4), indicating that the observed PTH(1–34)-mediated effects are mediated by the cAMP/PKA signaling cascade.

Time dependence of PTH(1–34) treatment on osteogenic and chondrogenic development

The time dependence of the opposite characteristics of PTH(1–34) treatment on osteo-/chondrogenic development was investigated in C3H10T½BMP2-PTHR or C3H10T½BMP2 cells. These cells were plated in the presence of PTH(1–34) (100 nM) at a density of 1000 cells/cm2. At distinct time points during the cultivation, PTH(1–34)-containing medium was replaced with fresh medium without PTH(1–34) (Fig. 7). Histological analysis was performed at day 10 postconfluence by assessing the number of alkaline positive, osteoblast-like cells (Fig. 7a) or the number of Alcian-Blue positive, chondrocyte-like cells (Fig. 7b). Osteoblast as well as chondrocyte development is stimulated when PTH(1–34) was present in the early stages of cultivation and suppressed if PTH(1–34) was present up to the final stages of the cultivation.

Figure FIG. 7.

Influence of a time-dependent PTH(1–34)-treatment on osteogenic and chondrogenic development. C3H10T½BMP2-PTHR or C3H10T½BMP2 cells were grown in the presence of PTH(1–34) (100 nM) at a starting plating density of 1000 cells/cm2. At the indicated time intervals, the medium was replaced with fresh medium without added PTH(1–34). All cells were cultivated until day 10 postconfluence. Then, histological analysis was performed by assessing (a) the number of Alcian-Blue positive, chondrocyte-like cells or (b) the number of alkaline positive, osteoblast-like cells. Values represent a typical experiment in triplicates.

DISCUSSION

C3H10T½ cells represent a relatively early stage of mesenchymal cell determination with the ability to differentiate into myoblasts, adipocytes, osteoblasts, and chondrocytes. Their multipotential nature and their responsiveness toward TGF-β and BMP treatment make this line a useful model system to explore the involvement of factors in various mesenchymal differentiation processes. In C3H10T½ cells, BMP initiates development into the osteogenic and chondrogenic lineage. Among the factors that in vivo modulate cellular osteogenic and chondrogenic development are PTH and PTHrP. The latter factor has recently been postulated to play an Indian Hedgehog-dependent role in the regulation of chondrogenic development by restricting the access of maturing chondrocytes for hypertrophy.9 In the system presented here, such a transition into hypertrophy has not been observed on the basis of collagen (X) expression (data not shown); however, PTH-dependent effects upon chondrogenic development involving also the formation of morphologically distinct chondrocytes could readily be detected and evaluated. Thereby, PTH(1–84) is the predominant secretory product of the parathyroid gland, but fragmentation of this intact peptide readily occurs in several peripheral organs.33 The amino-terminal PTH fragment has been demonstrated in numerous studies to be necessary and sufficient for activating the PTH/PTHrP receptor. Other PTH fragments, such as the midregional and the carboxy-terminal fragment, have also been suggested to exert biological functions.34,35 However, binding and activation of the PTH/PTHrP receptor by the latter PTH fragments is not absolutely established. Recently, a new receptor type has been postulated for the carboxy-terminal fragment.36 The same group also has evidence of yet another type of PTH receptor,37 but so far an additional PTH receptor type is predominantly expressed in brain and pancreas.38 This increasingly complex PTH receptor situation may explain why only the amino-terminal PTH fragment (1–34) was able to influence significantly osteogenic and chondrogenic development in the present study involving the “classical” PTH/PTHrP receptor.

The response of osteoblastic cultures to PTH seems also to depend critically on whether they are primary cultures or cell lines, and especially the initial degree of differentiation.39 Furthermore, biological effects of PTH strictly correlate with the receptor density.40,41 The PTH/PTHrP receptors are transiently induced by BMP treatment in C3H10T½ cells (Fig. 1a).25,27 Here, the recombinant expression of the PTH/PTHrP receptor correlates already in the absence of added ligand with an enhanced development into the osteogenic or chondrogenic lineage. More and larger osteogenic and chondrogenic amplification centers are observed in C3H10T½BMP2-PTHR cells which may form mineralizing nodules eventually (Fig. 2). This may be attributed to trace amounts of PTH or PTHrP in the medium. However, a substantial PTHrP synthesis by the recombinant C3H10T½ cells seems unlikely since PTHrP mRNA has only marginally been detected by RT-PCR in parental and recombinant C3H10T½ cells (data not shown). However, other studies involving the recombinant expression of seven-transmembrane receptors indicate that a high-level expression may result in a ligand-independent activation of signaling.42 Moreover, such an overexpression does not interfere with the activity of other G-coupled receptors nor disrupt the major signaling cascades.41–43 Such a weak ligand-independent activation of PTH/PTHrP receptors could be responsible for the increased rates in osteo-/chondrogenic development in C3H10T½BMP2-PTHR cells while the permanent PTH(1–34)-dependent activation of PTH/PTHrP receptors in C3H10T½BMP2-PTHR cells stimulates early and suppresses late stages of osteo-/chondrogenic development, substantiating the view that the opposite effects of PTH treatment depend on the degree of differentiation (Figs. 5, 6, 7).

From above results, we conclude that PTH and the PTH/PTHrP receptor system are essential factors for osteogenic or chondrogenic development. They mediate an efficient developmental rate into these mesenchymal lineages but themselves are not sufficient to determine development into the osteogenic or chondrogenic lineage. So, the recombinant PTH/PTHrP receptor in C3H10T½ cells was not able to mediate an efficient development into the osteogenic or chondrogenic lineage without the presence of enhanced rates of BMP expression (data not shown).

Why does the natural ligand PTH and, probably, PTHrP produce these opposite effects? One might think of PTH exerting different functions during osteo-/chondrogenic development: an early stimulatory function regarding the amplification of committed mesenchymal precursor cells and a repressing function where activated high levels of PTH/PTHrP receptors interfere with the morphological transition of determined mesenchymal progenitors into osteoblasts or chondrocytes at later stages of development. In vitro we could demonstrate with C3H10T½BMP2-PTHR cells both effects, but also several in vivo investigations are consistent with this view. Animals with a homogeneous deletion of the PTHrP gene develop osseous malformation most likely caused by abnormal proliferation and differentiation of growth plate chondrocytes into hypertrophic chondrocytes, indicating a regulatory role of PTHrP upon chondrocytic maturation.44 A similar situation has been obtained in mice harboring a homogeneous deletion in the PTH/PTHrP receptor gene.8 Recently, it has been suggested that PTHrP limits the numbers of these chondrocytes to enter hypertrophy and by this controls rates of endochondral bone formation.9 Another report describes a patient with a mutation resulting in a constitutively active PTH/PTHrP receptor exhibiting a permanent ligand-independent cAMP synthesis. Besides abnormalities in the mineral homeostasis, this patient suffered from a rare form of short-limbed dwarfism (Jensen-type metaphyseal chondrodysplasia).45 A constitutive active PTH/PTHrP receptor might be compared with our in vitro system where the permanent presence of PTH(1–34) constitutively activates the PTH/PTHrP receptor in C3H10T½BMP2-PTHR cells leading to a suppression of chondrocytic development. In addition, recent in vitro studies in other cellular systems support our studies, suggesting a role for PTH and its receptor in stimulating early and suppressing late stages of osteo-/chondrogenic development.46,47

Ancillary